Sleeping posture pressure image identification method based on HOG characteristic and machine learning

Abstract

The invention relates to a sleeping posture pressure image identification method based on an HOG characteristic and machine learning. The method is characterized by comprising the following steps of performing data acquisition, namely acquiring real-time pressure data which are obtained through detection when a user functions on a large-area flexible pressure sensor array mattress; performing image conversion, namely converting the real-time pressure data which are acquired in the first step to a pressure image, wherein the step comprises procedures of establishing the image in which the image coordinate is same with sensor array distribution, and converting the pressure number which is acquired on each sensor to the gray scale of the pixel on the corresponding image coordinate, thereby obtaining the pressure image which reflects pressure distribution on the sensor array; performing image pre-processing; performing image HOG characteristic extraction, namely performing HOG characteristic extraction on the pressure image which is pre-processed in the third step, thereby obtaining an HOG characteristic set of the sleeping gesture pressure image; and performing sleeping gesture identification based on machine learning.

Description

technical field [0001] The present invention relates to an image recognition method, in particular to a method for automatically recognizing sleeping postures, in particular to a sleeping posture pressure image recognition method based on HOG features and machine learning. Background technique [0002] With the continuous development of digital image technology, target detection is the most important and basic step in an intelligent system. In response to the current enthusiasm for research on sleeping positions, people's health issues are also receiving more and more attention. People spend one-third of their life in sleep. The quality of sleep is related to people's mental and physical health. Sleep quality Poor people are prone to nervousness, fatigue, inability to concentrate, or eating disorders. Determining indicators of sleep quality, such as sleep stages, sleep difficulties, and sleep positions, is critical for medical diagnosis. According to research, a good sleep...

AI Technical Summary

Problems solved by technology

Nakajima K, Matsumoto Y, Tamura T.A monitor for posture changes and respiration in bed using real time image sequence analysis[C] / / Engineering in Medicine and Biology Society, 2000. Proceedings of the 22nd Annual International Conference of the IEEE .IEEE,2000,1:51-54.) proposed a system based on visual signals to analyze sleep respiration and posture changes, but the low light at night brings a lot of noise to imaging, and visible light video imaging will bring serious privacy issues , very detrimental to the health treatment of hospital patients

Images